https://liliputing.com/2018/04/intel-launches-8th-gen-u-series-chips-with-iri...In the face of these multiple major customer based announcements to bypass Intel completely Intel cannot remain silent and the big brown vaporware cannons are loaded up with freshly thought up
8th Generation brown stinky bovine blatt, all still being done at a chunky 14nm size.
Core i7-8559U: 4 cores/8 threads, 2.7 GHz base/4.5 GHz boost frequency, 8MB cache
Core i5-8269U: 4 cores/8 threads, 2.6 GHz base/4.2 GHz boost frequency, 6MB cache
Core i5-8259U: 4 cores/8 threads, 2.3 GHz base/3.8 GHz boost frequency, 6MB cache
Core i3-8109U: 2 cores/4 threads, 3 GHz base/3.6 GHz boost frequency, 4MB cache
The new chips are all 28 watt processors with a new Mobile Intel 300 series chipset with upgraded I/O, integrated 802.11ac WiFi (for up to twice the throughput), and integrated USB 3.1 Gen2 support for data transfer speeds up to 10 Gb/s.
All told, these chips should be more energy-efficient than their 45 watt Coffee Lake-H or Kaby Lake-G cousins, which makes them more suitable for thin-and-light laptops and 2-in-1 tablets. But they should offer significantly better graphics performance than Intel’s 15 watt Kaby Lake Refresh chips when it comes to gaming, multimedia, photo and video editing, and other tasks that leverage a GPU.Read between the lines --- this is the 8th time Intel has trotted out a re-shuffled deck of 14nm design names with little to no real changes, a fact that becomes clear when they are benchmarked by the snarling frustrated chip benchmark reviewers.
But wait, the very next day something NEW is brought out and touted as being under development ......Intel brings hexa-core chips to laptops with Coffee Lake-H Here’s a run-down of the first Coffee Lake-H chips expected to ship:
Core i9-8950HK: 6 cores/12 threads, 2.9 GHz base/4.8 GHz boost frequency, 12MB cache, unlocked
Xeon E-2186M: 6 cores/12 threads, 2.9 GHz base/4.8 GHz boost frequency, 12MB cache
Xeon E-2176M: 6 cores/12 threads, 2.7 GHz base/4.4 GHz boost frequency, 12MB cache
Intel Core i7-8850H: 6 cores/12 threads, 2.6 GHz base/4.3 GHz boost frequency, 9MP cache, partially unlocked
Intel Core i7-8750H: 6 cores/12 threads, 2.2 GHz base/4.2 GHz boost frequency, 9MB cache
Intel Core i5-8400H: 4 cores/8 threads, 2.5 GHz base/4.2 GHz boost frequency, 8MB cache
Intel Core i5-8300H: 4 cores/8 threads, 2.3 GHz base/4 GHz boost frequency, 8MB cacheOK, it is all still at 14nm, with lots of new chipset names and a copying of an old trick, auto-sorting the chipsets and burning off two sets of core traces for the bottom two lowest performing core sets out of an 8 core product line to make up a new hexa core product that has never existed before as there weren't enough 8 core desktop products being made to have enough 6 core leftovers to sell as separate "laptop processors".
Weeeeee ....... such progress. And not a word about any fixes for Meltdown, Spectre or CTS, not even a sniff about any hardware fixes to recover the up to 30% real performance hit that has been lost inside the last year.
The new chips are based on the same 14nm++ architecture as Intel’s Coffee Lake chips for desktops. DUH, because that is what they are, leftover accumulated first run failures, sorted then truncated desktop and Xenon rackset chipsets that are being sold at a discount for what they can actually do.
These 45 watt processors offer more performance and consume more power than the 15 watt, quad-core Kaby Lake Refresh chips Intel launched last year. no shite But the company says despite their differences, both are part of its 8th-gen Core processor lineup. Go figure.
Basically Kaby Lake-R chips are U-series processors aimed at thin and light laptops. Coffee Lake-H processors are high-performance chips aimed at gaming laptops, mobile workstations, and other high-end systems.
Intel says these chips bring faster load times and higher frame rates for gamers, faster photo and video editing, and better support for virtual reality, among other things.
Some chips also include a new Intel Thermal Velocity Boost (TVB) feature which automatically bumps up the CPU boost clock speed under the right circumstances.OK, let's talk about the art of throttling overheating processors and this "Some chips also include a new Intel Thermal Velocity Boost (TVB) feature which automatically bumps up the CPU boost clock speed under the right circumstances" totally lying little "it never happens" BS ditty.
When the chipsets get reviewed it will become obvious that they turn in their maximum clock speeds when dead cold but within 10 seconds the chipsets heat up enough to require significant throttling action from the nanny chip.
But by implying it works ass backwards on demand this situation becomes a UNIQUE FEATURE that nobody else has.
Sad thing is that Rita and Joe are dumb enough to believe it is a real feature. And yes, it can happen, I guess, if the chip had gotten hot enough to TOTALLY ENGAGE the throttling brakes, then the chip actually cools off a smidgen from fan and heat sink actions then it could actually get dialed back up speedwise a little bit as it cools back down towards room temperature.
Some, anyway. Mebbe. But not under any real operating load, no, not ever. Real loads means real heat and results in a one way reaction to more and more throttling.
Bad thing here is
these newer generation chipsets are So Extremely Sensitive to heat buildup that the nanny chipset is set to be VERY conservative and shuts them down a whole lot when they just get warm just a little bit.
This is part of the reason that benchmarkers are saying the new generations of Intel products have had no actual single core throughput improvement over previous generations (going back to over 5 years ago). It is all nanny chip games nowadays playing off the excess core count to always roll the heaviest work load over to the coolest cores. The new multi-multi core chipsets mebbe draw a little less wall socket power doing this dance, but they don't really go any faster because they throttle and switch so much more.
Tacking on even more cores and even more treads really doesn't help all that much once going past 4 cores, actually, as the cooling dance becomes inefficient.Also, mind you I thought that none of it is real anyway, it was all "planning announcements".......... Boy, was I wrong. The Intel product announcements (some vapor, some not) are rolling out nonstop as we speak. THERE ARE STILL NO ANNOUNCEMENTS ABOUT ANY REAL HARDWARE FIXES FOR MELTDOWN, SPECTRE OR ANY OF THE CTS MESSES but them builders sure are a building for this Christmas season. But, I sure was right about the pend up demand thing, though.
https://liliputing.com/2018/04/origin-gives-evo-15-s-gaming-laptop-a-hexa-cor...https://liliputing.com/2018/04/msi-gs65-stealth-thin-is-a-0-7-inch-4-1-pound-...https://liliputing.com/2018/04/gigabyte-aero-gaming-laptops-get-coffee-lake-h...https://liliputing.com/2018/04/dells-4-pound-xps-15-laptop-now-supports-up-to...https://liliputing.com/2018/04/acer-launches-nitro-5-gaming-laptop-with-6-cor...https://liliputing.com/2018/04/asus-launches-crazy-powerful-and-pricey-rog-ze...https://liliputing.com/2018/04/samsung-launches-notebook-odyssey-z-gaming-lap...===================================================
Two weeks further along --- Intel has created a bit of a mess for themselves with their effort to dump off a stored stockpile of sorted, truncated 8 core chipsets, selling them off cheaply as a six core chipset.
The idea grew legs and took off, now poor Intel has to go DESIGN a six core chipset that they can make at that lower price point and still show a profit as demand has outstripped the pile of old accumulated sorting failures.
This tells me that Intel has gotten yet another message that their internal cost burdens are too high and
to simply survive they have to be able to sell that built on purpose six core at that price lower point. Same thought applies to all their other chipsets, get the internal costs way way down and actually be competitive, or lose it all to ARM.